Summary Research Project on Wikipt
- Brazil Catalyst Project
- Research Project on Wikipt
Summary
- August 2011 - Annoucement of the research
- September, first - Creation of page Portuguese Wikipedia Trends and Behavior
- September, 6th - Presentation of the pourpose of the project (to qualitative researcher) and a test task: The purpose of this project is to develop a narrative about Portuguese Wikipedia editor retention, in order to determine some critical areas that need attention from the community in order to continue to attract and retain great editors. To accomplish this, we will be asking a set of focused questions about onwiki editor interactions each week, which the qualitative researcher will answer based on close reading of contribution and revision histories of pt.wiki pages.
- The test task focused on three main points: Choose what you think is the most active community discussion page in Wiki.pt; Who is :doing most of the talking; What are the main topics of discussion; What is the tone of these discussions.
- September, 9 th -The beggining (answer to test task and first topics analyzed): We pointed out the ptwikipedia community has three main tension focus: the Village Pump pages (Announcements, Proposals and General); Requests for administrators and AfD´s. We chose to analyze the page of proposals in the Village Pump due to the number of active discussion and the impact these discussions main have on the dynamic of the community specially on the way it deals with new editors and take decisions. See http://pt.wikipedia.org/wiki/Wikip%C3%A9dia:Esplanada/propostas.
- Another important topic in discussion mentioned was related to the inability of the community to enforce the policies of the :project when so many decisions are taken by vote. (The link is Wikipédia:Esplanada/propostas/Voto_qualificado_em_PE_(28ago2011)) :Onptwiki, the AfD are decided by votes and a constant claim is that voters do not bind to policies and recomendations, using :their votes to game the system. In one side some editor claim more juridical security, i.e. the polices should be enforced on :AfDs, in this case by disregarding votes which do not present justifications based on the polices. In the other side some :editors see the "one person, one vote" system as the only legitimated form of democracy and deny any special responsability to :administrators. The tune is slightly more violent and involves the some type of editors (mainly veteran editors).
- September, 14th - Creation on Collab the page Community Diagnostics/ Social Structure Hypothesis http://collab.wikimedia.org/wiki/Community_Diagnostics/Social_Structure_Hypothesis#Arbcom_.28Conselho_de_Arbitragem.29
- This was the mainly page of the research by mid-october. The research was intense concearning to establish a timeline of policies :and overview of positions inside ptwiki community. The purpose was to better understand who holds the power within the community. :Some questions wich are not answered in this page specially on hypothesis table such as the statement (hypothesis): "Admins are :generally viewed favorably in the community, and it is a great honor to become one" was discussed after and the data tend to show :the role of admins. are not as valued as we believed what is already a problem inside community.
- Another work front in September was collecting some data concearning the request for adminship, use of tools like huggle and use :of bots as well as participation in community. We observed that decisions requires extensive involvement with the pt.wiki (e.g. :300 valid edits in the main namespace to have righ to vote).
http://meta.wikimedia.org/wiki/Research:Portuguese_Wikipedia_trends_and_behavior/Social_Structure
- September, 29th - contributions on Collab beginned to be moved to meta "Portuguese Wikipedia Trends and Behavior"
- September, 30th - Creation of the page General Editor Trends. We worked on this page throughout october specially collecting data related to the hypotheses and next steps
- October, 8th - Wikisampa: Presentation of the research for the Wikipt community. Exchange of experiences and views on community.
- In October (mid october onwards) the focus of the research was directed to the newcomer experience
- http://meta.wikimedia.org/wiki/Research:Portuguese_Wikipedia_trends_and_behavior/New_Editor_Experience . The templates were :analysed as well as how hard is becoming a new editor. CAPTCHA test absorbed a great part of our attention.
- November - Newcomer experience continued to retain our attention, it was the time for revision, take some conclusions about the :work as a whole and deepen on newcomer experience.
- http://meta.wikimedia.org/wiki/Research:Portuguese_Wikipedia_trends_and_behavior/Suggestions
Conclusions edit
Portuguese wikipedia has a long track history of disputes and difficulty to solve problems and take binding decisions; discussions are seldom driven towards consensus. Although community seems to be quieter now, some holdovers are still present, for instance they have not yet been able to elect checkusers and arbcom members.
Concearning to elections inside community we observed participation in wikipt decisions requires extensive involviment: hundreds of editions in the main namespace to have right to vote This.is not the norm of other language projects and it can be a barrier to new well-intentioned editors.
In spite of the perception of decreasing, no quantitative evidence was found to support the hypotheses that edits have been diminishing. The researcher thinks a deeper analyses must be carried out in order to better understand the effect of changings in the quality of edits. Perhaps a more frequently usage of templates and automatic tools may be hidding a reduction on edits. The number of administrators on portuguese wikipedia is low when compared to other language projects even when we count only the number of admins who fullfill the minimum requirement of activity on portuguese wikipediaWelcome templates were analysed and some of them were considered to long, aggressive and frightening, warmer templates were recomended since a previous research had shown on english wikipedia that the first messages a user received had an impact on the probability to stay.
CAPTCHA test was implemented as a provisory measure to curb vandalism and became a permanent feature. No evidence was found that CAPTCHA is efficient in fighting vandalism although it may be an important obstacle to retain new editors. Community has been suggested not use CAPTCHA to slow down human editors.
Finally it seems to be of great importance for community to understand why they have been unnable to define the role of each flag such as administrators, which would help them to find means of dispute resolution and taking decisions. A clearer social structure would make it easier to enforce local policies and mitigate the impact of trolling and sockpupetry, which would allow them to be more open and welcoming to newcomers and may be bring back good old editors.